pacman::p_load(sf, raster, spatstat, tmap, tidyverse)1st & 2nd Order Spatial Point Patterns Analysis Methods
3.1 Exercise Overview
In this hands-on exercise, I learn how to to analyze spatial point patterns in R. On top of that I learn to apply first- and second-order analyses to assess the randomness of point distributions, and how to visualize and interpret the spatial concentration of facilities using Kernel Density Estimation (KDE)
Spatial Point Pattern Analysis involves evaluating the distribution of a set of points on a surface. These points can represent the locations of various events or facilities, such as:
- Events: Crime occurrences, traffic accidents, or disease outbreaks.
- Facilities: Business services like coffee shops, fast food outlets, or essential facilities such as childcare and eldercare centers.
In this hands-on exercise, we aim to explore the spatial distribution of childcare centers in Singapore using functions from the spatstat package. Specifically, we will address the following questions:
- Randomness of Distribution: Are childcare centers in Singapore randomly distributed across the country?
- Clusters and Concentrations: If the distribution is not random, where are the areas with higher concentrations of childcare centers?
3.2 Data Acquisition
Three data set will be used to answer these question. They are:
- CHILDCARE: A point feature dataset containing the location and attribute information of childcare centers in Singapore. This dataset was downloaded from Data.gov.sg and is in GeoJSON format.
- MP14_SUBZONE_WEB_PL: A polygon feature dataset providing information on URA’s 2014 Master Plan Planning Subzone boundaries. This data is in ESRI Shapefile format and was also downloaded from Data.gov.sg.
- CostalOutline: A polygon feature dataset showing the national boundary of Singapore. This dataset is provided by the Singapore Land Authority (SLA) and is in ESRI Shapefile format.
3.3 Getting Started
For this exercise, we will use the following 5 R packages:
sf, a relatively new R package specially designed to import, manage and process vector-based geospatial data in R.
spatstat, a comprehensive package for point pattern analysis. We’ll use it to perform first- and second-order spatial point pattern analyses and to derive kernel density estimation (KDE) layers.
raster, a package for reading, writing, manipulating, and modeling gridded spatial data (rasters). We will use it to convert image outputs generated by spatstat into raster format.
maptools, a set of tools for manipulating geographic data, mainly used here to convert spatial objects into the ppp format required by spatstat.
tmap, a package for creating high-quality static and interactive maps, leveraging the Leaflet API for interactive visualizations.
To install and load these packages into the R environment, we use the p_load function from the pacman package:
3.4 Importing Data into R
3.4.1 Importing the spatial data
In this section, we’ll use the st_read() function from the sf package to import three geospatial datasets into R:
childcare_sf <- st_read("data/child-care-services-geojson.geojson") %>%
st_transform(crs = 3414)Reading layer `child-care-services-geojson' from data source
`C:\Users\blzll\OneDrive\Desktop\Y3S1\IS415\Quarto\IS415\Hands-on_Ex\data\child-care-services-geojson.geojson'
using driver `GeoJSON'
Simple feature collection with 1545 features and 2 fields
Geometry type: POINT
Dimension: XYZ
Bounding box: xmin: 103.6824 ymin: 1.248403 xmax: 103.9897 ymax: 1.462134
z_range: zmin: 0 zmax: 0
Geodetic CRS: WGS 84
sg_sf <- st_read(dsn = "data", layer="CostalOutline")Reading layer `CostalOutline' from data source
`C:\Users\blzll\OneDrive\Desktop\Y3S1\IS415\Quarto\IS415\Hands-on_Ex\data'
using driver `ESRI Shapefile'
Simple feature collection with 60 features and 4 fields
Geometry type: POLYGON
Dimension: XY
Bounding box: xmin: 2663.926 ymin: 16357.98 xmax: 56047.79 ymax: 50244.03
Projected CRS: SVY21
mpsz_sf <- st_read(dsn = "data",
layer = "MP14_SUBZONE_WEB_PL")Reading layer `MP14_SUBZONE_WEB_PL' from data source
`C:\Users\blzll\OneDrive\Desktop\Y3S1\IS415\Quarto\IS415\Hands-on_Ex\data'
using driver `ESRI Shapefile'
Simple feature collection with 323 features and 15 fields
Geometry type: MULTIPOLYGON
Dimension: XY
Bounding box: xmin: 2667.538 ymin: 15748.72 xmax: 56396.44 ymax: 50256.33
Projected CRS: SVY21
The following code chunk changes the referencing system of the newly created simple feature data frames to Singapore national projected coordinate system.
childcare_sf <- st_transform(childcare_sf, crs = 3414)
mpsz_sf <- st_transform(mpsz_sf, crs = 3414)3.4.2 Mapping the geospatial data sets
After verifying that all datasets share the same CRS, it’s useful to visualize them to confirm their spatial alignment.
tmap_mode("plot")
tm_shape(mpsz_sf) +
tm_polygons() +
tm_shape(childcare_sf) +
tm_dots() 
In this code chunk:
tm_shape(mpsz_sf): Sets the base layer to the Singapore subzones.tm_polygons(): Plots the subzones with a light blue fill and a dark blue border.tm_shape(childcare_sf): Adds an overlay layer for the childcare centers.tm_dots(): Plots the childcare centers as red dots with a black border.
tmap_mode('view')
tm_shape(childcare_sf) +
tm_dots()tmap_mode('plot')By ensuring that all geospatial layers align correctly within the same map extent, we confirm that their CRS and coordinate values are consistent—a critical aspect of any geospatial analysis.
tmap_mode('plot')Notice that at the interactive mode, tmap is using leaflet for R API. The advantage of this interactive pin map is it allows us to navigate and zoom around the map freely. We can also query the information of each simple feature (i.e. the point) by clicking of them. Last but not least, you can also change the background of the internet map layer. Currently, three internet map layers are provided. They are: ESRI.WorldGrayCanvas, OpenStreetMap, and ESRI.WorldTopoMap. The default is ESRI.WorldGrayCanvas.
3.5 Geospatial Data wrangling
Although simple feature data frame is gaining popularity again sp’s Spatial* classes, there are, however, many geospatial analysis packages require the input geospatial data in sp’s Spatial* classes. In this section, you will learn how to convert simple feature data frame to sp’s Spatial* class.
3.5.1 Converting sf data frames to sp’s Spatial* class
To work with certain geospatial analysis packages that require data in sp’s Spatial* class, you can convert simple feature (sf) data frames to these classes using the as_Spatial() function from the sf package. Below is the code to convert the sf objects to sp’s Spatial* classes:
childcare <- as_Spatial(childcare_sf)
mpsz <- as_Spatial(mpsz_sf)
sg <- as_Spatial(sg_sf)After conversion, you can display the information of these three Spatial* classes as follows:
childcareclass : SpatialPointsDataFrame
features : 1545
extent : 11203.01, 45404.24, 25667.6, 49300.88 (xmin, xmax, ymin, ymax)
crs : +proj=tmerc +lat_0=1.36666666666667 +lon_0=103.833333333333 +k=1 +x_0=28001.642 +y_0=38744.572 +ellps=WGS84 +towgs84=0,0,0,0,0,0,0 +units=m +no_defs
variables : 2
names : Name, Description
min values : kml_1, <center><table><tr><th colspan='2' align='center'><em>Attributes</em></th></tr><tr bgcolor="#E3E3F3"> <th>ADDRESSBLOCKHOUSENUMBER</th> <td></td> </tr><tr bgcolor=""> <th>ADDRESSBUILDINGNAME</th> <td></td> </tr><tr bgcolor="#E3E3F3"> <th>ADDRESSPOSTALCODE</th> <td>018989</td> </tr><tr bgcolor=""> <th>ADDRESSSTREETNAME</th> <td>1, MARINA BOULEVARD, #B1 - 01, ONE MARINA BOULEVARD, SINGAPORE 018989</td> </tr><tr bgcolor="#E3E3F3"> <th>ADDRESSTYPE</th> <td></td> </tr><tr bgcolor=""> <th>DESCRIPTION</th> <td></td> </tr><tr bgcolor="#E3E3F3"> <th>HYPERLINK</th> <td></td> </tr><tr bgcolor=""> <th>LANDXADDRESSPOINT</th> <td>0</td> </tr><tr bgcolor="#E3E3F3"> <th>LANDYADDRESSPOINT</th> <td>0</td> </tr><tr bgcolor=""> <th>NAME</th> <td>THE LITTLE SKOOL-HOUSE INTERNATIONAL PTE. LTD.</td> </tr><tr bgcolor="#E3E3F3"> <th>PHOTOURL</th> <td></td> </tr><tr bgcolor=""> <th>ADDRESSFLOORNUMBER</th> <td></td> </tr><tr bgcolor="#E3E3F3"> <th>INC_CRC</th> <td>08F73931F4A691F4</td> </tr><tr bgcolor=""> <th>FMEL_UPD_D</th> <td>20200826094036</td> </tr><tr bgcolor="#E3E3F3"> <th>ADDRESSUNITNUMBER</th> <td></td> </tr></table></center>
max values : kml_999, <center><table><tr><th colspan='2' align='center'><em>Attributes</em></th></tr><tr bgcolor="#E3E3F3"> <th>ADDRESSBLOCKHOUSENUMBER</th> <td></td> </tr><tr bgcolor=""> <th>ADDRESSBUILDINGNAME</th> <td></td> </tr><tr bgcolor="#E3E3F3"> <th>ADDRESSPOSTALCODE</th> <td>829646</td> </tr><tr bgcolor=""> <th>ADDRESSSTREETNAME</th> <td>200, PONGGOL SEVENTEENTH AVENUE, SINGAPORE 829646</td> </tr><tr bgcolor="#E3E3F3"> <th>ADDRESSTYPE</th> <td></td> </tr><tr bgcolor=""> <th>DESCRIPTION</th> <td>Child Care Services</td> </tr><tr bgcolor="#E3E3F3"> <th>HYPERLINK</th> <td></td> </tr><tr bgcolor=""> <th>LANDXADDRESSPOINT</th> <td>0</td> </tr><tr bgcolor="#E3E3F3"> <th>LANDYADDRESSPOINT</th> <td>0</td> </tr><tr bgcolor=""> <th>NAME</th> <td>RAFFLES KIDZ @ PUNGGOL PTE LTD</td> </tr><tr bgcolor="#E3E3F3"> <th>PHOTOURL</th> <td></td> </tr><tr bgcolor=""> <th>ADDRESSFLOORNUMBER</th> <td></td> </tr><tr bgcolor="#E3E3F3"> <th>INC_CRC</th> <td>379D017BF244B0FA</td> </tr><tr bgcolor=""> <th>FMEL_UPD_D</th> <td>20200826094036</td> </tr><tr bgcolor="#E3E3F3"> <th>ADDRESSUNITNUMBER</th> <td></td> </tr></table></center>
mpszclass : SpatialPolygonsDataFrame
features : 323
extent : 2667.538, 56396.44, 15748.72, 50256.33 (xmin, xmax, ymin, ymax)
crs : +proj=tmerc +lat_0=1.36666666666667 +lon_0=103.833333333333 +k=1 +x_0=28001.642 +y_0=38744.572 +ellps=WGS84 +towgs84=0,0,0,0,0,0,0 +units=m +no_defs
variables : 15
names : OBJECTID, SUBZONE_NO, SUBZONE_N, SUBZONE_C, CA_IND, PLN_AREA_N, PLN_AREA_C, REGION_N, REGION_C, INC_CRC, FMEL_UPD_D, X_ADDR, Y_ADDR, SHAPE_Leng, SHAPE_Area
min values : 1, 1, ADMIRALTY, AMSZ01, N, ANG MO KIO, AM, CENTRAL REGION, CR, 00F5E30B5C9B7AD8, 16409, 5092.8949, 19579.069, 871.554887798, 39437.9352703
max values : 323, 17, YUNNAN, YSSZ09, Y, YISHUN, YS, WEST REGION, WR, FFCCF172717C2EAF, 16409, 50424.7923, 49552.7904, 68083.9364708, 69748298.792
sgclass : SpatialPolygonsDataFrame
features : 60
extent : 2663.926, 56047.79, 16357.98, 50244.03 (xmin, xmax, ymin, ymax)
crs : +proj=tmerc +lat_0=1.36666666666667 +lon_0=103.833333333333 +k=1 +x_0=28001.642 +y_0=38744.572 +datum=WGS84 +units=m +no_defs
variables : 4
names : GDO_GID, MSLINK, MAPID, COSTAL_NAM
min values : 1, 1, 0, ISLAND LINK
max values : 60, 67, 0, SINGAPORE - MAIN ISLAND
3.5.2 Converting the Spatial* class into generic sp format
To prepare data for use in the spatstat package, you first need to convert the Spatial* classes into a generic sp format:
childcare_sp <- as(childcare, "SpatialPoints")
sg_sp <- as(sg, "SpatialPolygons")Then, you can display the properties of these sp objects:
childcare_spclass : SpatialPoints
features : 1545
extent : 11203.01, 45404.24, 25667.6, 49300.88 (xmin, xmax, ymin, ymax)
crs : +proj=tmerc +lat_0=1.36666666666667 +lon_0=103.833333333333 +k=1 +x_0=28001.642 +y_0=38744.572 +ellps=WGS84 +towgs84=0,0,0,0,0,0,0 +units=m +no_defs
sg_spclass : SpatialPolygons
features : 60
extent : 2663.926, 56047.79, 16357.98, 50244.03 (xmin, xmax, ymin, ymax)
crs : +proj=tmerc +lat_0=1.36666666666667 +lon_0=103.833333333333 +k=1 +x_0=28001.642 +y_0=38744.572 +datum=WGS84 +units=m +no_defs
3.5.3 Converting the generic sp format into spatstat’s ppp format
To analyze spatial point patterns, you need to convert the sp objects into ppp objects, which are used by the spatstat package:
childcare_ppp <- as.ppp(st_coordinates(childcare_sf), st_bbox(childcare_sf))
childcare_pppMarked planar point pattern: 1545 points
marks are numeric, of storage type 'double'
window: rectangle = [11203.01, 45404.24] x [25667.6, 49300.88] units
Now, let us plot childcare_ppp and examine the different.
plot(childcare_ppp)
You can take a quick look at the summary statistics of the newly created ppp object by using the code chunk below.
summary(childcare_ppp)Marked planar point pattern: 1545 points
Average intensity 1.91145e-06 points per square unit
*Pattern contains duplicated points*
Coordinates are given to 11 decimal places
marks are numeric, of type 'double'
Summary:
Min. 1st Qu. Median Mean 3rd Qu. Max.
0 0 0 0 0 0
Window: rectangle = [11203.01, 45404.24] x [25667.6, 49300.88] units
(34200 x 23630 units)
Window area = 808287000 square units
Notice the warning message about duplicates. In spatial point patterns analysis an issue of significant is the presence of duplicates. The statistical methodology used for spatial point patterns processes is based largely on the assumption that process are simple, that is, that the points cannot be coincident.
3.5.4 Handling duplicated points
To determine the existence of duplicate points, we can check for duplicates in your Point Pattern (PPP) object object by using the code chunk below.
any(duplicated(childcare_ppp))[1] TRUE
To count the number of points at each location, we can use the multiplicity() function as shown in the code chunk below.
multiplicity(childcare_ppp) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1 1 1 3 1 1 1 1 2 1 1 1 1 1 1 1
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
1 1 1 1 1 1 1 1 1 1 9 1 1 1 1 1
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64
1 1 1 1 1 1 2 1 1 3 1 1 1 1 1 1
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
1 1 1 1 1 2 1 1 1 1 1 2 1 1 1 1
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96
1 1 1 3 1 1 1 1 1 1 1 1 1 1 1 1
97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112
1 1 1 1 1 1 1 1 2 1 1 1 1 1 1 1
113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128
1 1 1 1 1 1 2 1 1 1 3 1 1 1 2 1
129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144
1 1 1 1 1 2 1 1 1 1 1 1 1 1 3 2
145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160
1 2 1 1 1 2 2 3 1 5 1 5 1 1 1 2
161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176
1 1 1 1 2 1 1 1 1 1 1 2 1 1 1 1
177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192
1 4 1 1 1 1 1 1 1 1 1 1 1 1 1 1
193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208
1 1 1 1 1 2 2 1 1 1 1 2 1 4 1 1
209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224
2 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1
225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256
1 1 1 1 2 1 1 1 1 1 1 1 1 1 1 1
257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272
1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 3
273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288
1 1 1 1 1 1 3 1 1 1 1 1 1 1 1 1
289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
1 1 1 1 1 1 1 9 1 1 2 1 1 1 1 1
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336
1 1 1 5 1 1 1 1 1 2 1 1 2 2 1 1
337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352
1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 1
353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368
1 1 1 1 9 1 1 1 1 1 1 1 1 1 1 1
369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384
1 3 1 1 1 1 1 1 1 1 1 1 1 1 1 1
385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416
1 1 2 1 1 1 1 1 1 1 2 1 1 1 1 1
417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432
1 1 1 1 1 1 1 2 1 1 2 1 1 1 1 1
433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448
1 1 1 1 2 1 1 1 1 1 1 1 1 1 1 1
449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464
1 1 9 9 1 1 1 1 1 1 1 1 1 1 2 1
465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480
2 1 1 1 1 1 1 1 1 1 1 1 2 2 1 1
481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512
1 1 1 1 1 1 2 1 1 1 1 1 1 1 1 2
513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528
1 1 1 1 1 1 1 1 1 1 1 2 1 1 3 1
529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560
1 1 1 1 1 1 1 1 1 3 1 1 1 1 1 1
561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576
2 2 2 1 1 1 1 2 1 1 2 1 1 1 2 1
577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592
1 2 1 1 1 1 1 9 1 4 1 2 1 1 1 1
593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608
2 1 1 1 1 1 1 1 2 1 2 1 1 1 1 1
609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624
1 1 1 1 1 1 1 1 1 2 1 2 1 1 1 1
625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4
657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672
1 1 1 1 1 1 1 3 1 1 1 1 1 1 1 1
673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688
1 1 1 1 1 4 1 1 1 1 1 4 1 1 1 1
689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720
1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1
721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752
1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1
753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768
1 1 1 1 1 2 1 1 1 1 1 1 1 1 1 1
769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784
1 1 1 1 1 1 1 1 1 4 1 1 1 1 1 1
785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848
1 1 1 1 1 1 1 2 1 1 1 1 1 1 1 1
849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2
881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896
3 1 1 1 2 1 1 1 3 1 1 3 1 1 1 1
897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2
961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056
1 1 1 1 1 1 1 1 1 2 2 1 1 1 1 1
1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120
1 1 1 1 1 2 1 1 1 1 1 1 1 1 1 1
1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136
1 1 1 1 1 1 1 1 2 2 1 1 1 5 1 1
1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152
1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 1
1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168
1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1
1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184
1 9 1 2 2 1 1 1 2 1 1 1 1 1 1 1
1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200
1 1 1 1 2 1 1 1 3 1 1 1 1 1 1 1
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216
9 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232
1 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1
1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2
1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312
1 1 1 2 1 2 1 1 1 2 2 2 1 1 1 1
1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328
1 1 2 1 1 1 1 1 1 1 1 1 2 1 1 1
1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344
1 1 1 1 3 1 1 1 1 1 1 1 1 1 1 1
1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360
1 1 1 1 1 1 1 1 4 1 1 1 1 1 2 1
1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392
1 1 1 1 1 1 1 1 1 9 1 1 1 1 1 1
1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440
1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1
1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456
1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472
1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1
1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488
1 1 1 1 1 1 2 1 1 1 1 1 1 1 1 1
1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504
1 1 1 1 1 1 1 1 1 1 5 1 1 1 1 1
1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536
1 1 1 1 1 2 1 1 1 1 2 1 1 1 1 3
1537 1538 1539 1540 1541 1542 1543 1544 1545
1 1 1 1 1 1 2 1 1
To check for the number of locations that have more than one point event, we can use the code chunk below.
sum(multiplicity(childcare_ppp) > 1)[1] 128
The output show 128 duplicated point events.
To view the locations of these duplicate point events, we will plot childcare data by using the code chunk below.
tmap_mode('view')
tm_shape(childcare) +
tm_dots(alpha=0.4,
size=0.05)tmap_mode('plot')
tmap_mode('view')There are three ways to overcome this problem.
The first and easiest way is to delete the duplicates. But, that will also mean that some useful point events will be lost.
The second solution is use jittering. If duplicates are hard to spot, you can apply a slight jitter to the points’ coordinates. Jittering will slightly displace the points so that overlapping points are separated on the map.
The third solution is to make each point “unique” and then attach the duplicates of the points to the patterns as marks, as attributes of the points. Then you would need analytical techniques that take into account these marks.
The jitter parameter will slightly move each point by a small, random amount. This can help to visually separate points that are in the same space.
tm_shape(childcare) +
tm_dots(jitter=0.1, alpha=0.4, size=0.05)childcare_ppp_jit <- rjitter(childcare_ppp,
retry=TRUE,
nsim=1,
drop=TRUE)After jittering we can check if there are any duplicate point in this geospatial data
any(duplicated(childcare_ppp_jit))[1] FALSE
3.5.5 Creating owin object
To confine analysis to a geographical area, convert the SpatialPolygon object to an owin object of spatstat:
sg_owin <- as.owin(sg_sf)
plot(sg_owin)
Further analysis can be done through the summary() function of Base R:
summary(sg_owin)Window: polygonal boundary
50 separate polygons (1 hole)
vertices area relative.area
polygon 1 (hole) 30 -7081.18 -9.76e-06
polygon 2 55 82537.90 1.14e-04
polygon 3 90 415092.00 5.72e-04
polygon 4 49 16698.60 2.30e-05
polygon 5 38 24249.20 3.34e-05
polygon 6 976 23344700.00 3.22e-02
polygon 7 721 1927950.00 2.66e-03
polygon 8 1992 9992170.00 1.38e-02
polygon 9 330 1118960.00 1.54e-03
polygon 10 175 925904.00 1.28e-03
polygon 11 115 928394.00 1.28e-03
polygon 12 24 6352.39 8.76e-06
polygon 13 190 202489.00 2.79e-04
polygon 14 37 10170.50 1.40e-05
polygon 15 25 16622.70 2.29e-05
polygon 16 10 2145.07 2.96e-06
polygon 17 66 16184.10 2.23e-05
polygon 18 5195 636837000.00 8.78e-01
polygon 19 76 312332.00 4.31e-04
polygon 20 627 31891300.00 4.40e-02
polygon 21 20 32842.00 4.53e-05
polygon 22 42 55831.70 7.70e-05
polygon 23 67 1313540.00 1.81e-03
polygon 24 734 4690930.00 6.47e-03
polygon 25 16 3194.60 4.40e-06
polygon 26 15 4872.96 6.72e-06
polygon 27 15 4464.20 6.15e-06
polygon 28 14 5466.74 7.54e-06
polygon 29 37 5261.94 7.25e-06
polygon 30 111 662927.00 9.14e-04
polygon 31 69 56313.40 7.76e-05
polygon 32 143 145139.00 2.00e-04
polygon 33 397 2488210.00 3.43e-03
polygon 34 90 115991.00 1.60e-04
polygon 35 98 62682.90 8.64e-05
polygon 36 165 338736.00 4.67e-04
polygon 37 130 94046.50 1.30e-04
polygon 38 93 430642.00 5.94e-04
polygon 39 16 2010.46 2.77e-06
polygon 40 415 3253840.00 4.49e-03
polygon 41 30 10838.20 1.49e-05
polygon 42 53 34400.30 4.74e-05
polygon 43 26 8347.58 1.15e-05
polygon 44 74 58223.40 8.03e-05
polygon 45 327 2169210.00 2.99e-03
polygon 46 177 467446.00 6.44e-04
polygon 47 46 699702.00 9.65e-04
polygon 48 6 16841.00 2.32e-05
polygon 49 13 70087.30 9.66e-05
polygon 50 4 9459.63 1.30e-05
enclosing rectangle: [2663.93, 56047.79] x [16357.98, 50244.03] units
(53380 x 33890 units)
Window area = 725376000 square units
Fraction of frame area: 0.401
3.5.6 Combining point events object and owin object
Finally, you can combine the point events with the polygon feature to create a ppp object confined to the Singapore region
childcareSG_ppp = childcare_ppp[sg_owin]
summary(childcareSG_ppp )Marked planar point pattern: 1545 points
Average intensity 2.129929e-06 points per square unit
*Pattern contains duplicated points*
Coordinates are given to 11 decimal places
marks are numeric, of type 'double'
Summary:
Min. 1st Qu. Median Mean 3rd Qu. Max.
0 0 0 0 0 0
Window: polygonal boundary
50 separate polygons (1 hole)
vertices area relative.area
polygon 1 (hole) 30 -7081.18 -9.76e-06
polygon 2 55 82537.90 1.14e-04
polygon 3 90 415092.00 5.72e-04
polygon 4 49 16698.60 2.30e-05
polygon 5 38 24249.20 3.34e-05
polygon 6 976 23344700.00 3.22e-02
polygon 7 721 1927950.00 2.66e-03
polygon 8 1992 9992170.00 1.38e-02
polygon 9 330 1118960.00 1.54e-03
polygon 10 175 925904.00 1.28e-03
polygon 11 115 928394.00 1.28e-03
polygon 12 24 6352.39 8.76e-06
polygon 13 190 202489.00 2.79e-04
polygon 14 37 10170.50 1.40e-05
polygon 15 25 16622.70 2.29e-05
polygon 16 10 2145.07 2.96e-06
polygon 17 66 16184.10 2.23e-05
polygon 18 5195 636837000.00 8.78e-01
polygon 19 76 312332.00 4.31e-04
polygon 20 627 31891300.00 4.40e-02
polygon 21 20 32842.00 4.53e-05
polygon 22 42 55831.70 7.70e-05
polygon 23 67 1313540.00 1.81e-03
polygon 24 734 4690930.00 6.47e-03
polygon 25 16 3194.60 4.40e-06
polygon 26 15 4872.96 6.72e-06
polygon 27 15 4464.20 6.15e-06
polygon 28 14 5466.74 7.54e-06
polygon 29 37 5261.94 7.25e-06
polygon 30 111 662927.00 9.14e-04
polygon 31 69 56313.40 7.76e-05
polygon 32 143 145139.00 2.00e-04
polygon 33 397 2488210.00 3.43e-03
polygon 34 90 115991.00 1.60e-04
polygon 35 98 62682.90 8.64e-05
polygon 36 165 338736.00 4.67e-04
polygon 37 130 94046.50 1.30e-04
polygon 38 93 430642.00 5.94e-04
polygon 39 16 2010.46 2.77e-06
polygon 40 415 3253840.00 4.49e-03
polygon 41 30 10838.20 1.49e-05
polygon 42 53 34400.30 4.74e-05
polygon 43 26 8347.58 1.15e-05
polygon 44 74 58223.40 8.03e-05
polygon 45 327 2169210.00 2.99e-03
polygon 46 177 467446.00 6.44e-04
polygon 47 46 699702.00 9.65e-04
polygon 48 6 16841.00 2.32e-05
polygon 49 13 70087.30 9.66e-05
polygon 50 4 9459.63 1.30e-05
enclosing rectangle: [2663.93, 56047.79] x [16357.98, 50244.03] units
(53380 x 33890 units)
Window area = 725376000 square units
Fraction of frame area: 0.401
plot(childcareSG_ppp)
3.6 First-order Spatial Point Patterns Analysis
In this section, we will learn how to perform first-order Spatial Point Pattern Analysis (SPPA) using the spatstat package. The focus will be on:
Deriving Kernel Density Estimation (KDE) layers for visualizing and exploring the intensity of point processes.
Performing Confirmatory Spatial Point Patterns Analysis using Nearest Neighbour statistics.
3.6.1 Kernel Density Estimation
Kernel Density Estimation (KDE) is a non-parametric way to estimate the intensity (density) of spatial point patterns. It helps in visualizing the spatial distribution of points by smoothing the point pattern to create a continuous surface.
3.6.1.1 Computing kernel density estimation using automatic bandwidth selection method
To compute the KDE for the spatial point pattern of childcare services in Singapore, we’ll use the density() function from the spatstat package. This function allows for various configurations:
Automatic Bandwidth Selection: We’ll use bw.diggle() , a method that selects an optimal bandwidth based on the data. Other methods like bw.CvL() , bw.scott(), or bw.ppl() can also be used depending on the specific needs of the analysis.
Smoothing Kernel: The default kernel used is Gaussian. Other options include “Epanechnikov”, “Quartic”, or “Disc”.
Edge Correction: The intensity estimate is corrected for edge effects to reduce bias, following methods described by Jones (1993) and Diggle (2010).
kde_childcareSG_bw <- density(childcareSG_ppp,
sigma=bw.diggle,
edge=TRUE,
kernel="gaussian")
plot(kde_childcareSG_bw)
In this example, the density values range from 0 to 0.000035, which are quite small. This is because the unit of measurement is in meters, making the density values “number of points per square meter”.
To check the bandwidth used:
bw <- bw.diggle(childcareSG_ppp)
bw sigma
298.4095
3.6.1.2 Rescalling KDE values
The small density values are due to the measurement unit being in meters. To make the results more interpretable, we can rescale the spatial point pattern from meters to kilometers.
The code chunk below rescale the point pattern to kilometers, recompute the KDE with the rescaled data, and plots the rescaled KDE
childcareSG_ppp.km <- rescale.ppp(childcareSG_ppp, 1000, "km")
kde_childcareSG_bw <- density(childcareSG_ppp.km,
sigma=bw.diggle,
edge=TRUE,
kernel="gaussian")
plot(kde_childcareSG_bw)
The KDE output will look identical to the original, but the density values will now be more comprehensible, reflecting “number of points per square kilometer.”
3.6.2 Working with different automatic badwidth methods
Different bandwidth selection methods can produce different smoothing results:
bw.CvL(): Cross-validation based on the likelihood.bw.scott(): Scott’s rule of thumb.bw.ppl(): Likelihood cross-validation proposed by Diggle.
bw.CvL(childcareSG_ppp.km) sigma
4.543278
bw.scott(childcareSG_ppp.km) sigma.x sigma.y
2.224898 1.450966
bw.ppl(childcareSG_ppp.km) sigma
0.3897114
bw.diggle(childcareSG_ppp.km) sigma
0.2984095
Recommendation: 1. Baddeley et. (2016) suggested the use of the bw.ppl() algorithm because in ther experience it tends to produce the more appropriate values when the pattern consists predominantly of tight clusters.
- But they also insist that if the purpose of once study is to detect a single tight cluster in the midst of random noise then the bw.diggle() method seems to work best.
kde_childcareSG.ppl <- density(childcareSG_ppp.km,
sigma=bw.ppl,
edge=TRUE,
kernel="gaussian")
childcareSG_ppp.km <- rescale.ppp(childcareSG_ppp, 1000, "km")
kde_childcareSG.bw <- density(childcareSG_ppp.km,
sigma=bw.diggle,
edge=TRUE,
kernel="gaussian")
par(mfrow=c(1,2))
plot(kde_childcareSG.bw, main = "bw.diggle")
plot(kde_childcareSG.ppl, main = "bw.ppl")
3.6.3 Working with different kernel methods
Beyond the Gaussian kernel, three other kernels can be used to compute KDE:
Epanechnikov
Quartic
Disc
par(mfrow=c(2,2), mar=c(1, 1, 1, 1), cex=0.5)
plot(density(childcareSG_ppp.km,
sigma=bw.ppl,
edge=TRUE,
kernel="gaussian"),
main="Gaussian")
plot(density(childcareSG_ppp.km,
sigma=bw.ppl,
edge=TRUE,
kernel="epanechnikov"),
main="Epanechnikov")
plot(density(childcareSG_ppp.km,
sigma=bw.ppl,
edge=TRUE,
kernel="quartic"),
main="Quartic")
plot(density(childcareSG_ppp.km,
sigma=bw.ppl,
edge=TRUE,
kernel="disc"),
main="Disc")
3.7 Fixed and Adaptive KDE
3.7.1 Computing KDE by using fixed bandwidth
Next, you will compute a KDE layer by defining a bandwidth of 600 meter. Notice that in the code chunk below, the sigma value used is 0.6. This is because the unit of measurement of childcareSG_ppp.km object is in kilometer, hence the 600m is 0.6km.
To compute a KDE layer using a fixed bandwidth of 0.6 km, use the following code:
kde_childcareSG_600 <- density(childcareSG_ppp.km, sigma=0.6, edge=TRUE, kernel="gaussian")
plot(kde_childcareSG_600)
This will generate a KDE layer with a consistent bandwidth across the entire study area, useful for uniform spatial point patterns.
3.7.2 Computing KDE by using adaptive bandwidth
Adaptive bandwidth methods are more suitable for spatial point patterns with high variability, such as urban versus rural areas. The adaptive.density() function of spatstat can be used to create a KDE layer that adjusts the bandwidth based on point density:
kde_childcareSG_adaptive <- adaptive.density(childcareSG_ppp.km, method="kernel")
plot(kde_childcareSG_adaptive)
To compare the outputs of fixed and adaptive bandwidth KDE:
par(mfrow=c(1,2))
plot(kde_childcareSG.bw, main = "Fixed Bandwidth")
plot(kde_childcareSG_adaptive, main = "Adaptive Bandwidth")
3.7.3 Converting KDE output into grid object.
For mapping purposes, KDE outputs can be converted into grid and raster formats 1. Converting to Grid Object:
gridded_kde_childcareSG_bw <- as(kde_childcareSG.bw, "SpatialGridDataFrame")
spplot(gridded_kde_childcareSG_bw)
- Converting to Raster Object:
kde_childcareSG_bw_raster <- raster(kde_childcareSG.bw)
kde_childcareSG_bw_rasterclass : RasterLayer
dimensions : 128, 128, 16384 (nrow, ncol, ncell)
resolution : 0.4170614, 0.2647348 (x, y)
extent : 2.663926, 56.04779, 16.35798, 50.24403 (xmin, xmax, ymin, ymax)
crs : NA
source : memory
names : layer
values : -8.476185e-15, 28.51831 (min, max)
- Assigning Projection System: Ensure the CRS (Coordinate Reference System) is assigned correctly:
projection(kde_childcareSG_bw_raster) <- CRS("+init=EPSG:3414")
kde_childcareSG_bw_rasterclass : RasterLayer
dimensions : 128, 128, 16384 (nrow, ncol, ncell)
resolution : 0.4170614, 0.2647348 (x, y)
extent : 2.663926, 56.04779, 16.35798, 50.24403 (xmin, xmax, ymin, ymax)
crs : +proj=tmerc +lat_0=1.36666666666667 +lon_0=103.833333333333 +k=1 +x_0=28001.642 +y_0=38744.572 +ellps=WGS84 +units=m +no_defs
source : memory
names : layer
values : -8.476185e-15, 28.51831 (min, max)
- Visualizing the Raster: Finally, use the
tmappackage to visualize the raster in a cartographic map:
tm_shape(kde_childcareSG_bw_raster) +
tm_raster("layer", palette = "viridis") +
tm_layout(legend.position = c("right", "bottom"), frame = FALSE)Notice that the raster values are encoded explicitly onto the raster pixel using the values in “v”” field.
3.7.4 Comparing Spatial Point Patterns using KDE
In this section, we will learn to compare KDE of childcare at Ponggol, Tampines, Chua Chu Kang and Jurong West planning regions.
3.7.4.1 Extracting study area
pg <- mpsz_sf %>%
filter(PLN_AREA_N == "PUNGGOL")
tm <- mpsz_sf %>%
filter(PLN_AREA_N == "TAMPINES")
ck <- mpsz_sf %>%
filter(PLN_AREA_N == "CHOA CHU KANG")
jw <- mpsz_sf %>%
filter(PLN_AREA_N == "JURONG WEST")Plotting target planning areas 1. Ponggol
par(mfrow=c(2,2))
plot(pg, main = "Ponggol")
- Tampines
plot(tm, main = "Tampines")
- Choa Chu Kang
plot(ck, main = "Choa Chu Kang")
- Jurong West
plot(jw, main = "Jurong West")
3.7.4.2 Creating owin object
Now, we will convert these sf objects into owin objects that is required by spatstat.
The owin objects represent the study areas as window objects, which are necessary for spatial point pattern analysis in spatstat. These are created by converting the sf objects (pg, tm, ck, and jw) representing different regions into owin format:
pg_owin = as.owin(pg)
tm_owin = as.owin(tm)
ck_owin = as.owin(ck)
jw_owin = as.owin(jw)3.7.4.3 Combining childcare points and the study area
we are then able to extract childcare that is within the specific region to perform analysis later on.
childcare_pg_ppp = childcare_ppp_jit[pg_owin]
childcare_tm_ppp = childcare_ppp_jit[tm_owin]
childcare_ck_ppp = childcare_ppp_jit[ck_owin]
childcare_jw_ppp = childcare_ppp_jit[jw_owin]The childcare centers within each specific region are extracted using the owin objects.
These point patterns (ppp objects) are then rescaled from meters to kilometers:
childcare_pg_ppp.km = rescale.ppp(childcare_pg_ppp, 1000, "km")
childcare_tm_ppp.km = rescale.ppp(childcare_tm_ppp, 1000, "km")
childcare_ck_ppp.km = rescale.ppp(childcare_ck_ppp, 1000, "km")
childcare_jw_ppp.km = rescale.ppp(childcare_jw_ppp, 1000, "km")Finally, the four study areas and the locations of the childcare centers are plotted:
par(mfrow=c(2,2), mar=c(1, 1, 1, 1), cex=0.5)
plot(childcare_pg_ppp.km, main="Punggol")
plot(childcare_tm_ppp.km, main="Tampines")
plot(childcare_ck_ppp.km, main="Choa Chu Kang")
plot(childcare_jw_ppp.km, main="Jurong West")
3.7.4.4 Computing KDE
The Kernel Density Estimate (KDE) for each area is computed using the density() function, with the bw.diggle method to derive the bandwidth:
par(mfrow=c(2,2), mar=c(1, 1, 1, 1), cex=0.5)
plot(density(childcare_pg_ppp.km, sigma=bw.diggle, edge=TRUE, kernel="gaussian"), main="Punggol")
plot(density(childcare_tm_ppp.km, sigma=bw.diggle, edge=TRUE, kernel="gaussian"), main="Tampines")
plot(density(childcare_ck_ppp.km, sigma=bw.diggle, edge=TRUE, kernel="gaussian"), main="Choa Chu Kang")
plot(density(childcare_jw_ppp.km, sigma=bw.diggle, edge=TRUE, kernel="gaussian"), main="Jurong West")
3.7.4.5 Computing Fixed Bandwidth KDE
For comparison, a fixed bandwidth of 250 meters is used to compute KDE for the same areas:
par(mfrow=c(2,2), mar=c(1, 1, 1, 1), cex=0.5)
plot(density(childcare_pg_ppp.km, sigma=0.25, edge=TRUE, kernel="gaussian"), main="Punggol")
plot(density(childcare_tm_ppp.km, sigma=0.25, edge=TRUE, kernel="gaussian"), main="Tampines")
plot(density(childcare_ck_ppp.km, sigma=0.25, edge=TRUE, kernel="gaussian"), main="Choa Chu Kang")
plot(density(childcare_jw_ppp.km, sigma=0.25, edge=TRUE, kernel="gaussian"), main="Jurong West")
3.8 Nearest Neighbour Analysis
In this section, we will perform the Clark-Evans test of aggregation for a spatial point pattern by using clarkevans.test() of statspat.
3.8.1 Testing Spatial Point Patterns using Clark and Evans Test
The Clark-Evans test is performed to assess the spatial distribution of the childcare centers. The hypotheses are:
H0: The distribution of childcare centers is random.
H1: The distribution of childcare centers is clustered.
The test is conducted as follows:
clarkevans.test(childcareSG_ppp, correction="none", clipregion="sg_owin", alternative="clustered", nsim=99)
Clark-Evans test
No edge correction
Z-test
data: childcareSG_ppp
R = 0.55631, p-value < 2.2e-16
alternative hypothesis: clustered (R < 1)
Result Interpretation:
R = 0.55631: The R-value is less than 1, indicating a tendency towards clustering.
p-value < 2.2e-16: The p-value is extremely small, suggesting that the clustering pattern is statistically significant. The null hypothesis of CSR (Complete Spatial Randomness) is rejected in favor of the alternative hypothesis, indicating that the childcare centers in Singapore are clustered.
3.8.2 Clark and Evans Test: Punggol planning area
In the code chunk below, clarkevans.test() of spatstat is used to performs Clark-Evans test of aggregation for childcare centre in Punggol planning area.
clarkevans.test(childcare_pg_ppp,
correction="none",
clipregion=NULL,
alternative="two.sided",
nsim=999)
Clark-Evans test
No edge correction
Z-test
data: childcare_pg_ppp
R = 0.93275, p-value = 0.315
alternative hypothesis: two-sided
Interpretation:
- R = 0.91163: The R-value is close to 1, indicating a spatial distribution that is close to random.
- p-value = 0.1867: The p-value is greater than 0.05, meaning there is no statistically significant evidence to reject the null hypothesis of CSR. This suggests that the childcare centers in the Punggol area are randomly distributed and do not exhibit significant clustering or regularity.
3.9 Second-order Spatial Point Patterns Analysis
This section introduces second-order analyses of spatial point patterns, focusing on measuring interaction between points.
3.10 Analysing Spatial Point Process Using G-Function
The G function measures the distribution of the distances from an arbitrary event to its nearest event. In this section, you will learn how to compute G-function estimation by using Gest() of spatstat package. You will also learn how to perform monta carlo simulation test using envelope() of spatstat package.
3.10.1 Choa Chu Kang planning area
3.10.1.1 Computing G-function estimation
The code chunk below is used to compute G-function using Gest() of spatat package.
G_CK = Gest(childcare_ck_ppp, correction = "border")
plot(G_CK, xlim=c(0,500))
3.10.1.2 Performing Complete Spatial Randomness Test
To perform a Complete Spatial Randomness (CSR) test using a Monte Carlo simulation with the G-function in R, you are correctly using the envelope() function from the spatstat package. Here’s how you can carry out the test and interpret the results:
Hypothesis Definition:
Null Hypothesis (Ho): The distribution of childcare services at Choa Chu Kang is randomly distributed (i.e., follows CSR).
Alternative Hypothesis (H1): The distribution of childcare services at Choa Chu Kang is not randomly distributed.
Set Up the Test:
- You will perform a Monte Carlo test using the G-function, which measures the distribution of nearest-neighbor distances.
G_CK.csr <- envelope(childcare_ck_ppp, Gest, nsim = 999)Generating 999 simulations of CSR ...
1, 2, 3, ......10.........20.........30.........40.........50.........60..
.......70.........80.........90.........100.........110.........120.........130
.........140.........150.........160.........170.........180.........190........
.200.........210.........220.........230.........240.........250.........260......
...270.........280.........290.........300.........310.........320.........330....
.....340.........350.........360.........370.........380.........390.........400..
.......410.........420.........430.........440.........450.........460.........470
.........480.........490.........500.........510.........520.........530........
.540.........550.........560.........570.........580.........590.........600......
...610.........620.........630.........640.........650.........660.........670....
.....680.........690.........700.........710.........720.........730.........740..
.......750.........760.........770.........780.........790.........800.........810
.........820.........830.........840.........850.........860.........870........
.880.........890.........900.........910.........920.........930.........940......
...950.........960.........970.........980.........990........
999.
Done.
plot(G_CK.csr)
3.10.2 Tampines planning area
3.10.2.1 Computing G-function estimation
G_tm = Gest(childcare_tm_ppp, correction = "best")
plot(G_tm)
3.10.2.2 Performing Complete Spatial Randomness Test
To confirm the observed spatial patterns above, we perform the Complete Spatial Randomness (CSR) test for the distribution of childcare services in Tampines using the G-function in R, you can follow the steps and code chunk below:
Hypothesis:
Null Hypothesis (Ho): The distribution of childcare services at Tampines is randomly distributed (CSR).
Alternative Hypothesis (H1): The distribution of childcare services at Tampines is not randomly distributed
G_tm.csr <- envelope(childcare_tm_ppp, Gest, correction = "all", nsim = 999)Generating 999 simulations of CSR ...
1, 2, 3, ......10.........20.........30.........40.........50.........60..
.......70.........80.........90.........100.........110.........120.........130
.........140.........150.........160.........170.........180.........190........
.200.........210.........220.........230.........240.........250.........260......
...270.........280.........290.........300.........310.........320.........330....
.....340.........350.........360.........370.........380.........390.........400..
.......410.........420.........430.........440.........450.........460.........470
.........480.........490.........500.........510.........520.........530........
.540.........550.........560.........570.........580.........590.........600......
...610.........620.........630.........640.........650.........660.........670....
.....680.........690.........700.........710.........720.........730.........740..
.......750.........760.........770.........780.........790.........800.........810
.........820.........830.........840.........850.........860.........870........
.880.........890.........900.........910.........920.........930.........940......
...950.........960.........970.........980.........990........
999.
Done.
plot(G_tm.csr)
3.11 Analysing Spatial Point Process Using F-Function
The F-function estimates the empty space function F(r) from a point pattern within a defined window. It provides insights into the spatial distribution by measuring the distribution of distances from a randomly chosen location in the study area to the nearest event (e.g., childcare centers). We will compute the F-function for the Choa Chu Kang and Tampines planning areas and perform a Complete Spatial Randomness (CSR) test using Monte Carlo simulations.
We will learn how to compute F-function estimation by using Fest() of spatstat package, and how to perform monta carlo simulation test using envelope() of spatstat package.
3.11.1 Choa Chu Kang planning area
3.11.1.1 Computing F-fucntion estimate
F_CK = Fest(childcare_ck_ppp)
plot(F_CK)
3.11.1.2 Performing Complete Spatial Randomness Test
To confirm the observed spatial patterns at Choa Chu Kang, a hypothesis test using the F-function will be conducted. The hypotheses are:
Ho (Null Hypothesis): The distribution of childcare services at Choa Chu Kang is randomly distributed.
H1 (Alternative Hypothesis): The distribution of childcare services at Choa Chu Kang is not randomly distributed.
The null hypothesis will be rejected if the p-value is smaller than the alpha value of 0.001.
Monte Carlo Test Using F-function:
F_CK.csr <- envelope(childcare_ck_ppp, Fest, nsim = 999)Generating 999 simulations of CSR ...
1, 2, 3, ......10.........20.........30.........40.........50.........60..
.......70.........80.........90.........100.........110.........120.........130
.........140.........150.........160.........170.........180.........190........
.200.........210.........220.........230.........240.........250.........260......
...270.........280.........290.........300.........310.........320.........330....
.....340.........350.........360.........370.........380.........390.........400..
.......410.........420.........430.........440.........450.........460.........470
.........480.........490.........500.........510.........520.........530........
.540.........550.........560.........570.........580.........590.........600......
...610.........620.........630.........640.........650.........660.........670....
.....680.........690.........700.........710.........720.........730.........740..
.......750.........760.........770.........780.........790.........800.........810
.........820.........830.........840.........850.........860.........870........
.880.........890.........900.........910.........920.........930.........940......
...950.........960.........970.........980.........990........
999.
Done.
plot(F_CK.csr)
3.11.2 Tampines planning area
3.11.2.1 Computing F-fucntion estimation
Monte Carlo test with F-fucntion
F_tm = Fest(childcare_tm_ppp, correction = "best")
plot(F_tm)
3.11.2.2 Computing F-fucntion estimation
Similar to before, a hypothesis test will be conducted to confirm the observed spatial patterns above. The hypothesis and test are as follows:
Ho = The distribution of childcare services at Tampines are randomly distributed.
H1= The distribution of childcare services at Tampines are not randomly distributed.
The null hypothesis will be rejected if p-value is smaller than alpha value of 0.001.
The code chunk below is used to perform the hypothesis testing.
F_tm.csr <- envelope(childcare_tm_ppp, Fest, correction = "all", nsim = 999)Generating 999 simulations of CSR ...
1, 2, 3, ......10.........20.........30.........40.........50.........60..
.......70.........80.........90.........100.........110.........120.........130
.........140.........150.........160.........170.........180.........190........
.200.........210.........220.........230.........240.........250.........260......
...270.........280.........290.........300.........310.........320.........330....
.....340.........350.........360.........370.........380.........390.........400..
.......410.........420.........430.........440.........450.........460.........470
.........480.........490.........500.........510.........520.........530........
.540.........550.........560.........570.........580.........590.........600......
...610.........620.........630.........640.........650.........660.........670....
.....680.........690.........700.........710.........720.........730.........740..
.......750.........760.........770.........780.........790.........800.........810
.........820.........830.........840.........850.........860.........870........
.880.........890.........900.........910.........920.........930.........940......
...950.........960.........970.........980.........990........
999.
Done.
plot(F_tm.csr)
3.12 Analysing Spatial Point Process Using K-Function
K-function measures the number of events found up to a given distance of any particular event. In this section, you will learn how to compute K-function estimates by using Kest() of spatstat package. You will also learn how to perform monta carlo simulation test using envelope() of spatstat package.
3.12.1 Choa Chu Kang planning area
3.12.1.1 Computing K-fucntion estimation
K_ck = Kest(childcare_ck_ppp, correction = "Ripley")
plot(K_ck, . -r ~ r, ylab= "K(d)-r", xlab = "d(m)")
3.12.1.2 Performing Complete Spatial Randomness Test
Similar to before, a hypothesis test will be conducted to confirm the observed spatial patterns above. The hypothesis and test are as follows:
Ho = The distribution of childcare services at Choa Chu Kang are randomly distributed.
H1= The distribution of childcare services at Choa Chu Kang are not randomly distributed.
The null hypothesis will be rejected if p-value is smaller than alpha value of 0.001.
The code chunk below is used to perform the hypothesis testing.
K_ck.csr <- envelope(childcare_ck_ppp, Kest, nsim = 99, rank = 1, global=TRUE)Generating 99 simulations of CSR ...
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20,
21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60,
61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80,
81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98,
99.
Done.
plot(K_ck.csr, . - r ~ r, xlab="d", ylab="K(d)-r")
3.12.2 Tampines planning area
3.12.2.1 Computing K-fucntion estimation
K_tm = Kest(childcare_tm_ppp, correction = "Ripley")
plot(K_tm, . -r ~ r,
ylab= "K(d)-r", xlab = "d(m)",
xlim=c(0,1000))
3.12.2.2 Performing Complete Spatial Randomness Test
Similar to before, a hypothesis test will be conducted to confirm the observed spatial patterns above. The hypothesis and test are as follows:
Ho = The distribution of childcare services at Tampines are randomly distributed.
H1= The distribution of childcare services at Tampines are not randomly distributed.
The null hypothesis will be rejected if p-value is smaller than alpha value of 0.001.
The code chunk below is used to perform the hypothesis testing.
K_tm.csr <- envelope(childcare_tm_ppp, Kest, nsim = 99, rank = 1, global=TRUE)Generating 99 simulations of CSR ...
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20,
21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60,
61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80,
81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98,
99.
Done.
plot(K_tm.csr, . - r ~ r,
xlab="d", ylab="K(d)-r", xlim=c(0,500))
3.13 Analysing Spatial Point Process Using L-Function
The L-function is a method used to analyze spatial point patterns by transforming the K-function to be more interpretable, particularly by normalizing the function against a theoretical CSR model. This section covers how to compute the L-function estimation and perform a Monte Carlo simulation test using the Lest() and envelope() of spatstat package.
3.13.1 Choa Chu Kang planning area
3.13.1.1 Computing L-fucntion estimation
To compute the L-function estimation for Choa Chu Kang, use the Lest() function with the “Ripley” correction. Then, plot the results.
L_ck = Lest(childcare_ck_ppp, correction = "Ripley")
plot(L_ck, . -r ~ r,
ylab= "L(d)-r", xlab = "d(m)")
3.13.1.2 Performing Complete Spatial Randomness Test
Similar to before, a hypothesis test will be conducted to confirm the observed spatial patterns above. The hypothesis and test are as follows:
Ho = The distribution of childcare services at Tampines are randomly distributed.
H1= The distribution of childcare services at Tampines are not randomly distributed.
The null hypothesis will be rejected if p-value is smaller than alpha value of 0.001.
The code chunk below is used to perform the hypothesis testing.
L_ck.csr <- envelope(childcare_ck_ppp, Lest, nsim = 99, rank = 1, global=TRUE)Generating 99 simulations of CSR ...
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20,
21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60,
61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80,
81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98,
99.
Done.
plot(L_ck.csr, . - r ~ r, xlab="d", ylab="L(d)-r")
3.13.2 Tampines planning area
3.13.2.1 Computing L-fucntion estimation
L_tm = Lest(childcare_tm_ppp, correction = "Ripley")
plot(L_tm, . -r ~ r,
ylab= "L(d)-r", xlab = "d(m)",
xlim=c(0,1000))
3.13.2.2 Performing Complete Spatial Randomness Test
Similar to before, a hypothesis test will be conducted to confirm the observed spatial patterns above. The hypothesis and test are as follows:
Ho = The distribution of childcare services at Tampines are randomly distributed.
H1= The distribution of childcare services at Tampines are not randomly distributed.
The null hypothesis will be rejected if p-value is smaller than alpha value of 0.001.
The code chunk below is used to perform the hypothesis testing.
L_tm.csr <- envelope(childcare_tm_ppp, Lest, nsim = 99, rank = 1, global=TRUE)Generating 99 simulations of CSR ...
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20,
21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60,
61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80,
81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98,
99.
Done.
plot(L_tm.csr, . - r ~ r, xlab="d", ylab="L(d)-r", xlim=c(0,500))